[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[XEN v5 10/11] xen/Arm: GICv3: Define macros to read/write 64 bit


  • To: <xen-devel@xxxxxxxxxxxxxxxxxxxx>
  • From: Ayan Kumar Halder <ayan.kumar.halder@xxxxxxx>
  • Date: Mon, 5 Dec 2022 13:26:36 +0000
  • Arc-authentication-results: i=1; mx.microsoft.com 1; spf=pass (sender ip is 165.204.84.17) smtp.rcpttodomain=lists.xenproject.org smtp.mailfrom=amd.com; dmarc=pass (p=quarantine sp=quarantine pct=100) action=none header.from=amd.com; dkim=none (message not signed); arc=none
  • Arc-message-signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=microsoft.com; s=arcselector9901; h=From:Date:Subject:Message-ID:Content-Type:MIME-Version:X-MS-Exchange-AntiSpam-MessageData-ChunkCount:X-MS-Exchange-AntiSpam-MessageData-0:X-MS-Exchange-AntiSpam-MessageData-1; bh=d/f6KcnxjmuXzQKyqBzEdB2sXtJfFweqelRMZY8EzeE=; b=nIurm1d6qwmxIfuRCQ/BR6fQmtj06FPJILsjE/RONtqspp1xJPrLd0eyAerJSk5hJ0ZJAvAO7+PehIkM+xwsVJRvChUx1j6ayWf1E0z3mFW41Px5/KWwjHXsFhGwJ2tW4PyPCZ3uFh1zNyn0n/5sJhh+9c6yP2pCQKMFR1d3GFZK9QMr5i71LBR74VCFj+qQlxuIzdu2AER5W+Xdo8JEhBxKcfyoPAoLyeWnBOAq7sax/hJbAnAvK2Qt75EDquRzTtmYdB8DYDb4tWH8q1gw9Eli72wGYUe6lwqkaRCoZbeE4AMGl/iiXO8N8o8UIxtuWZG+BHnTIEUGcMsVcRu5vg==
  • Arc-seal: i=1; a=rsa-sha256; s=arcselector9901; d=microsoft.com; cv=none; b=ibNqC8FoPo7rMaFCj9d/D0xnm0rvzJjzhtx/VdM5AJsRntfmyRaYL3l0lqvbHtlf/DW/x/LlAzGzBflmLNK8Z09k0k0OKy2J1FjCXm9H2szuMuZ+LdC2Fj8Mop9hhagFq1tJEMR2QuRp4Zv9piMxV3OUk7U8RXyMAPFeq9BAjTYZ35jc+oUaggdofbe16SYLc4Tkk7zTfHNaJBDCQcR1FWiFJCW+s6H4sowhKDfCXh+/vcIRI/XzfeAijPT+YtAu3V1Ghr7MpMVL0xKjSogjqJkQg9WrCyWN754ljmXAD3mMUjevxxvmTlL3PwvWKgu+2JCltSo/cbSPZREBQZMTvQ==
  • Cc: <sstabellini@xxxxxxxxxx>, <stefanos@xxxxxxxxxx>, <julien@xxxxxxx>, <Volodymyr_Babchuk@xxxxxxxx>, <bertrand.marquis@xxxxxxx>, <michal.orzel@xxxxxxx>, <jgrall@xxxxxxxxxx>, <burzalodowa@xxxxxxxxx>, "Ayan Kumar Halder" <ayan.kumar.halder@xxxxxxx>
  • Delivery-date: Mon, 05 Dec 2022 14:53:37 +0000
  • List-id: Xen developer discussion <xen-devel.lists.xenproject.org>

On AArch32, ldrd/strd instructions are not atomic when used to access MMIO.
Furthermore, ldrd/strd instructions are not decoded by Arm when running as
a guest to access emulated MMIO region.
Thus, we have defined readq_relaxed_non_atomic()/writeq_relaxed_non_atomic()
which in turn calls readl_relaxed()/writel_relaxed() for the lower and upper
32 bits.
For AArch64, readq_relaxed_non_atomic()/writeq_relaxed_non_atomic() invokes
readq_relaxed()/writeq_relaxed() respectively.
As GICv3 registers (GICD_IROUTER, GICR_TYPER) can be accessed in a non atomic
manner, so we have used readq_relaxed_non_atomic()/readq_relaxed_non_atomic().

However, the following points are noted for the non atomic access :-
1. In gicv3_dist_init(), using non atomic write on GICD_IROUTER is fine as this
gets invoked when interrupts are disabled.
2. In gicv3_populate_rdist(), using non atomic read on GICR_TYPER is fine as
the register is read and the interrupts are disabled as well.
3. In gicv3_irq_set_affinity(), using non atomic write on GICD_IROUTER. This
may be called with interrupts enabled. So, a non-atomic access (on AArch32)
means the GIC will see a transient value when only one of two 32-bit will be
updated. However, only AFF3 is defined in the upper 32 bits and they are 0, so
this will never change.
On AArch64, writeq_relaxed_non_atomic() invokes writeq_relaxed() (which is
atomic), so this problem does not arise.

Signed-off-by: Ayan Kumar Halder <ayan.kumar.halder@xxxxxxx>
---

Changes from :-
v1 - 1. Use ldrd/strd for readq_relaxed()/writeq_relaxed().
2. No need to use le64_to_cpu() as the returned byte order is already in cpu
endianess.

v2 - 1. Replace {read/write}q_relaxed with {read/write}q_relaxed_non_atomic().

v3 - 1. Use inline function definitions for {read/write}q_relaxed_non_atomic().
2. For AArch64, {read/write}q_relaxed_non_atomic() should invoke 
{read/write}q_relaxed().
Thus, we can avoid any ifdef in gic-v3.c.

v4 - 1. Updated the commit message.

 xen/arch/arm/gic-v3.c               |  6 +++---
 xen/arch/arm/include/asm/arm32/io.h | 20 ++++++++++++++++++++
 xen/arch/arm/include/asm/arm64/io.h |  2 ++
 3 files changed, 25 insertions(+), 3 deletions(-)

diff --git a/xen/arch/arm/gic-v3.c b/xen/arch/arm/gic-v3.c
index 6457e7033c..3c5b88148c 100644
--- a/xen/arch/arm/gic-v3.c
+++ b/xen/arch/arm/gic-v3.c
@@ -651,7 +651,7 @@ static void __init gicv3_dist_init(void)
     affinity &= ~GICD_IROUTER_SPI_MODE_ANY;
 
     for ( i = NR_GIC_LOCAL_IRQS; i < nr_lines; i++ )
-        writeq_relaxed(affinity, GICD + GICD_IROUTER + i * 8);
+        writeq_relaxed_non_atomic(affinity, GICD + GICD_IROUTER + i * 8);
 }
 
 static int gicv3_enable_redist(void)
@@ -745,7 +745,7 @@ static int __init gicv3_populate_rdist(void)
         }
 
         do {
-            typer = readq_relaxed(ptr + GICR_TYPER);
+            typer = readq_relaxed_non_atomic(ptr + GICR_TYPER);
 
             if ( (typer >> 32) == aff )
             {
@@ -1265,7 +1265,7 @@ static void gicv3_irq_set_affinity(struct irq_desc *desc, 
const cpumask_t *mask)
     affinity &= ~GICD_IROUTER_SPI_MODE_ANY;
 
     if ( desc->irq >= NR_GIC_LOCAL_IRQS )
-        writeq_relaxed(affinity, (GICD + GICD_IROUTER + desc->irq * 8));
+        writeq_relaxed_non_atomic(affinity, (GICD + GICD_IROUTER + desc->irq * 
8));
 
     spin_unlock(&gicv3.lock);
 }
diff --git a/xen/arch/arm/include/asm/arm32/io.h 
b/xen/arch/arm/include/asm/arm32/io.h
index 73a879e9fb..782b564809 100644
--- a/xen/arch/arm/include/asm/arm32/io.h
+++ b/xen/arch/arm/include/asm/arm32/io.h
@@ -80,10 +80,30 @@ static inline u32 __raw_readl(const volatile void __iomem 
*addr)
                                         __raw_readw(c)); __r; })
 #define readl_relaxed(c) ({ u32 __r = le32_to_cpu((__force __le32) \
                                         __raw_readl(c)); __r; })
+/*
+ * ldrd instructions are not decoded by Arm when running as a guest to access
+ * emulated MMIO region. Thus, readq_relaxed_non_atomic() invokes 
readl_relaxed()
+ * twice to read the lower and upper 32 bits.
+ */
+static inline u64 readq_relaxed_non_atomic(const volatile void __iomem *addr)
+{
+        u64 val = (((u64)readl_relaxed(addr + 4)) << 32) | readl_relaxed(addr);
+        return val;
+}
 
 #define writeb_relaxed(v,c)     __raw_writeb(v,c)
 #define writew_relaxed(v,c)     __raw_writew((__force u16) cpu_to_le16(v),c)
 #define writel_relaxed(v,c)     __raw_writel((__force u32) cpu_to_le32(v),c)
+/*
+ * strd instructions are not decoded by Arm when running as a guest to access
+ * emulated MMIO region. Thus, writeq_relaxed_non_atomic() invokes 
writel_relaxed()
+ * twice to write the lower and upper 32 bits.
+ */
+static inline void writeq_relaxed_non_atomic(u64 val, volatile void __iomem 
*addr)
+{
+        writel_relaxed((u32)val, addr);
+        writel_relaxed((u32)(val >> 32), addr + 4);
+}
 
 #define readb(c)                ({ u8  __v = readb_relaxed(c); __iormb(); __v; 
})
 #define readw(c)                ({ u16 __v = readw_relaxed(c); __iormb(); __v; 
})
diff --git a/xen/arch/arm/include/asm/arm64/io.h 
b/xen/arch/arm/include/asm/arm64/io.h
index 30bfc78d9e..2e2ab24f78 100644
--- a/xen/arch/arm/include/asm/arm64/io.h
+++ b/xen/arch/arm/include/asm/arm64/io.h
@@ -102,11 +102,13 @@ static inline u64 __raw_readq(const volatile void __iomem 
*addr)
 #define readw_relaxed(c)        ({ u16 __v = le16_to_cpu((__force 
__le16)__raw_readw(c)); __v; })
 #define readl_relaxed(c)        ({ u32 __v = le32_to_cpu((__force 
__le32)__raw_readl(c)); __v; })
 #define readq_relaxed(c)        ({ u64 __v = le64_to_cpu((__force 
__le64)__raw_readq(c)); __v; })
+#define readq_relaxed_non_atomic(c)           readq_relaxed(c)
 
 #define writeb_relaxed(v,c)     ((void)__raw_writeb((v),(c)))
 #define writew_relaxed(v,c)     ((void)__raw_writew((__force 
u16)cpu_to_le16(v),(c)))
 #define writel_relaxed(v,c)     ((void)__raw_writel((__force 
u32)cpu_to_le32(v),(c)))
 #define writeq_relaxed(v,c)     ((void)__raw_writeq((__force 
u64)cpu_to_le64(v),(c)))
+#define writeq_relaxed_non_atomic(v,c)        writeq_relaxed(v,c)
 
 /*
  * I/O memory access primitives. Reads are ordered relative to any
-- 
2.17.1




 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.